New Scientist – June 10 2017

NEWS & TECHNOLOGY
Brain signals
used to recreate
photos of faces
SXS
Black hole merger
rattles the cosmos
Leah Crane
THREE’S a party. For the third
time, the LIGO collaboration has
detected gravitational waves
emanating from a pair of merging
black holes – yielding clues about
how these duos form and building
up our catalogue of them.
“The first one was a novelty.
The second one was confirmation
that the novelty of the first one
was not a fluke. The third one
is astrophysics,” says LIGO
spokesperson David Shoemaker
at the Massachusetts Institute of
Technology (MIT).
LIGO detects waveforms,
which are readouts of the ripples
in the fabric of space-time caused
by masses moving through it. The
spins of merging black holes can
warp those waveforms, which are
mostly produced by their orbits
and eventual collision.
The first event yielded too
little information to determine
the direction of each black hole’s
spin. The second provided a bit
more information, indicating
that each black hole was probably
spinning in the same direction as
they were orbiting one another.
But this third pair of black
holes tilts towards Earth in a
different way from the other
two, according to Shoemaker,
allowing LIGO to see more
about how each one spins.
This view has revealed that
they aren’t spinning in the same
direction as their orbit. That
means they’re probably spinning
“Spins, and particularly
misaligned spins, will help
us figure out how pairs of
merging black holes form”
in different directions or – far
less likely – not spinning at all.
“Spins, and particularly
misaligned spins, will help us
figure out how these things are
formed,” says Carl Rodriguez
at MIT. Going beyond detection
to examining these objects’
properties turns this into a “new
branch of astronomy”, he adds.
Black hole binaries are either
born together from a pair of
orbiting stars, or form separately
in a dense stellar cluster and later
drift together at its centre. In the
first case, the pair should rotate in
the same direction they orbit, as
binary stars do. In the second, says
Rodriguez, “they’re pointing in
whatever directions they please”.
LIGO’s second detection, a black
hole binary discovered in 2015,
seemed to be from black holes
born orbiting together. But this
new pair, found on 4 January,
may have formed independently.
At least one of the black holes
seems to spin in a different
direction to its orbit. The
differences indicate that both
formation scenarios can occur.
Because this new black hole
binary is about 3 billion light
years away – twice as far as the
others we’ve detected – its
gravitational waves have to
ripple through more space-time
before they reach Earth. That
distance allows us to get greater
insight into potential deviations
from Einstein’s theory of general
relativity (PhysicalReview
Letters, doi.org/b73r).
General relativity states that all
gravitational waves should travel
at the same speed – the speed of
light. Because the waves seemed
to do that in this case, even over
such a huge distance, they backed
up Einstein’s cosmic rule.
The research marks the start
of an era of using gravitational
waves to study the cosmic kin of
black hole binaries. ■
PRECISION images of real faces have
been recreated by monitoring the
activity of certain cells in the brains of
macaque monkeys as they looked at
photographs of people.
The study is the first to provide a
full and simple explanation of how
the brains of macaques – and by
implication, humans – generate
composite images of any face they
see. “We’ve cracked the brain’s code
for facial identity,” says Doris Tsao at
–All in a spin– the California Institute of Technology.
The brain has regions of specialised
face cells, which become active when
a person sees a face. Tsao and her
colleague, Steven Le Chang, inserted
electrodes into three patches of these
cells in macaques, enabling them to
record the activity of 205 neurons.
The pair then showed three of
these macaques 2000 images of
human faces. They discovered that
each of the face cells is tuned to view
faces in slightly different ways – as if
photographing a face from multiple
angles at once. The combined signals
from these cells encode 50 different
aspects of a face – for example,
shape, distance between eyes and
skin texture.
When all these are combined,
they give a clear composite image.
“The key is that even though there’s
an infinite number of faces, you can
describe all of them with just these
50 dimensions,” says Tsao.
The researchers developed
algorithms from the face-cell feedback
that enabled them to recreate
composite facial images from monkey
brain-cell activity (Cell, doi.org/b73v).
It is likely that memories of
familiar faces are held by a different
type of cell in the hippocampus.
“Tsao’s work provides the first specific
hypothesis for how the response of
face cells in the cortex can be utilised
by cells in the hippocampus to form
memories of individuals we’ve seen
before,” says Ueli Rutishauser at the
Cedars-Sinai Medical Center in Los
Angeles. Andy Coghlan ■
14|NewScientist|10June2017